Radiology peer review and peer learning software: Score the read. Share the lesson.
Review final reads in RADIN, record what needs attention, and turn useful cases into a de-identified teaching library for your organization.
Original images
Final report · impression
Small acute subdural hematoma along the left convexity, up to 4 mm. No midline shift. No other acute intracranial finding.
Final · the review is not blindPeer review
A thin subdural on a motion-degraded scan.
Category Great Call · Subspecialty Neuroradiology
Teaching point Check both convexities on every head CT, even when motion degrades the study.
3M+
Studies a yearProcessed in RADIN.
200+
RadiologistsRead in RADIN.
100
RadiologistsIn the largest group on RADIN.
12+
HospitalsRun on RADIN.
Peer review
Reviewed where the radiologist already reads.
Assign, review and save in the same screens your radiologists use every day.
-
Assigned
A user picks a final study, a reviewer and a review type.
-
On the worklist
The study moves to Assigned for peer review and reaches the reviewer's worklist, following your worklist filters.
-
Reviewed
The reviewer sees what the reader had: the images, the documents, notes and priors, and the final report.
-
Saved
The study returns to Final. The review is kept in its own peer review database, signed and timestamped.
Flag a case when a question comes in
When an ED or hospital physician raises a possible miss, a radiologist assigns the study for review from the worklist and picks the colleague who should look at it.
Correct the report from the review
When the report needs correcting, Create Addendum starts the addendum from the peer review panel, without leaving the case.
The score
Five scores, and what kind of discrepancy it was.
RADIN keeps the 1 to 3 peer review scale radiologists know, with a and b for clinical significance. From 2a up, the reviewer also says what kind of discrepancy it was. Nothing is preselected.
| Score | Meaning | Discrepancy type |
|---|---|---|
| 1 | Agrees with the original interpretation. | Not asked |
| 2a | A discrepancy most radiologists might reasonably make. Unlikely to be clinically significant. | Required |
| 2b | A discrepancy most radiologists might reasonably make. Likely to be clinically significant. | Required |
| 3a | A discrepancy that should be identified most of the time. Unlikely to be clinically significant. | Required |
| 3b | A discrepancy that should be identified most of the time. Likely to be clinically significant. | Required |
Not seen
The finding is visible on the images and not described in the report.
Seen but misinterpreted
The finding is described, but its significance or the diagnosis is wrong.
Overcall
The report describes a finding that is not there, or gives a normal finding undue weight.
Beyond the score
The read can be right and the report still wrong.
Every review answers three more questions, and can record a technical problem and a great call. The required answers open empty, and a review cannot be saved without them.
| Field | Values | How it works |
|---|---|---|
| Critical or actionable finding communication | Met · Not met · N/A | Required. The study's critical-results log (recipient, time, method) shows beside the field, so the reviewer sees what was documented without leaving the form. |
| Follow-up recommendation | Appropriate · Missing · Not guideline-concordant · N/A | Required. Not guideline-concordant covers the wrong interval, test or assessment category, or more or less follow-up than the guideline calls for. |
| Report accuracy and clarity | No issue · Minor · Significant | Required. Minor is style, phrasing or organization. Significant is an error that changes meaning: wrong side or level, a dictation error, a wrong measurement, an impression that contradicts the findings, or an available prior not compared. |
| System or technical issue | Image quality or technique · Missing history or priors · Workflow or transmission | Optional, with one line. It describes the study and how it was handled, not the radiologist, and it is never counted against the reader. |
| Great Call | One line | Optional. A subtle finding, or a read that changed management. A review can be all praise. |
Each answer can carry a short note, and the reviewer's comments stay free text. A clean review closes itself: anything that needs a second look goes to Pending confirmation for your peer review team.
Peer review tracking
Where every review stands, from score to committee.
Track due dates, responses and decisions in one place, visible only to the team managing the review.
- AssignedSystem
- Review completedSystem
- No further actionSystem
- Pending confirmationSystem · 5 days
- Confirmed
- Notice sent, awaiting response10 days
- Response received
- Case conference scheduled
- Decision pending
- Decision issued15 days
- Appeal filed
- Scheduled for committee
- Committee decision issued
- Closed
Shown as DisputedDefault due times in business days
Stages you can edit
RADIN sets the first four stages; your team sets the rest. Each organization edits the list, marks which stages end a review and sets a due time for each. Every change is logged, so the dashboard shows the days spent in each stage and every review past its due date.
Review type, set at assignment
- Random
- Focused, routine
- Focused, for cause
- Client request
- Complaint
- Discrepancy found on a later study
- Addendum review
- Critical-results audit
- Other
The reviewer never sees it. Random reviews are reported apart from all the others.
Outcome and flags
- Upheld
- Modified
- Vacated
- No action
A vacated review leaves every performance count, and the original values stay in the log. Flags record what followed: an addendum issued, a patient-safety follow-up, an educational conference, a focused review, external reporting and more.
Notes that cannot be rewritten
Internal notes carry their author and time. They cannot be edited or deleted, only struck through with a reason, and they never reach the reviewer, the reading radiologist or a teaching case.
For the reading radiologist
See the result. Never the reviewer.
Reviews of their own studies
With permission, radiologists can see the review results, comments and teaching points for studies they signed. The review date is shown as a day only, with a Disputed label when applicable.
Reviewer identity is removed before the review leaves the server.
Hidden
- The reviewer's name, initials, picture, signature and user ID
- The time of day, so the reviewer cannot be guessed from it
- The tracking stage, outcome, review type and internal notes
Peer learning
Every shared case, de-identified, in one library.
Shared cases reach the Peer learning tab through the workflow below. The original clinical study stays unchanged.
01 · Shared
From a review or a study
The radiologist saves. There is nothing else to do.
02 · Copied
A new study, de-identified
RADIN copies the study, and any priors chosen, under a case number in place of the patient.
03 · Checked
Every header audited
Clean cases are released within minutes. A case that may carry text in the pixels waits for your peer learning administrator.
04 · On the tab
Open to every radiologist
Filter by subspecialty, category and modality, or show only the cases you have not reviewed.
How a copy is de-identified
| Patient name, ID, birth date and other identifiers | Removed. The patient becomes the case number. Sex is kept, and age in years, with 90 and over shown as 90+. |
|---|---|
| Dates and times | Shifted by one random offset per case, the same for the study and its priors, so the intervals stay true. |
| Institution, devices, operators and physicians | Removed. |
| UIDs and accession number | Regenerated. |
| Private tags and overlays | Removed, following the DICOM basic de-identification profile. |
| Report and addenda | Copied without header, demographics, signature or radiologist, through a PHI scrubber. The teaching point goes through it too. |
| Paperwork, tech notes, documents and the order | Not copied. |
- Great Call2026-0142 · Splenic laceration, 25-year-old, motor vehicle collisionBody · CT · New · 31 of 58 reviewedNot reviewed
- Interpretation2026-0131 · Thin left convexity subdural hematomaNeuro · CT · 52 of 58 reviewed✓ Nov 10
Reviewed, one radiologist at a time
Mark reviewed records each radiologist separately, so a case can be new for one and done for another. Radiologists see the count; the administrator sees the names.
Teaching cases never touch the clinical work
A copy never appears on a clinical worklist or in RADIN SELECT, billing, turnaround counts, RADINA agents or client reports, and it is never routed, sent or exported.
Run by your administrator
Your peer learning administrator releases held cases, retries or discards a failed copy, changes a subspecialty and retires a case with a reason.
Dashboards, OPPE and FPPE
Counted from one log, so the numbers always agree.
Every assignment, review, share and stage change writes one entry to a log that cannot be edited. Every dashboard number is counted from it, for any date range, and can be reproduced from it, ready for OPPE and FPPE.
Each radiologist
Their own numbers: reviews completed, cases contributed, and, with peer learning, cases reviewed and those still to see.
Your peer review team
Every radiologist, shares by category and subspecialty, technical issues by type, clinic and modality, reviews by stage, outcome and review type, and the overdue list. Export as CSV.
Performance
Scores and the new fields as counts, with Great Calls received. Findings still pending confirmation are shown apart, vacated reviews are left out, and technical issues are never counted against the reader.
Peer review results also appear in RADIN Analytics, beside workload, turnaround and RVUs.
FAQ
Frequently asked questions
Can the reading radiologist be the reviewer?
No. RADIN will not assign a review to the radiologist who read the study. The review is not blind: the reviewer sees the final report with the images.
Does RADIN run our peer review committee?
No. Your team decides who screens cases, how the radiologist is notified and when the committee meets. RADIN records that process.
What happens to our existing reviews?
Reviews done before go-live come with you: completed ones are closed, open ones stay assigned, and the new fields show as Not recorded rather than counted.
Peer review & peer learning in RADIN™
What does your peer review leave out?
Bring the peer review form you use today. We'll show you the same review in RADIN, and the teaching case it could become.
RADIN™ is a trademark of RADIN, LLC. Screens marked Illustrative show example workflows with invented patient and study data.